Photovoltaic cells where light is converted into electric power have been prevailing in a wide range of application fields such as consumer electronics, industrial electronics, and space exploration. In consumer electronics, photovoltaic cells that consist of materials such as amorphous silicon are choices for a variety of inexpensive and low power applications. Typical conversion efficiency, i.e. the solar cell conversion efficiency, of amorphous silicon based photovoltaic cells ranges between ˜10% [Yamamoto K, Yoshimi M, Suzuki T, Tawada Y, Okamoto T, Nakajima A. Thin film poly-Si solar cell on glass substrate fabricated at low temperature. Presented at MRS Spring Meeting, San Francisco, April 1998.]. Although the fabrication processes of amorphous silicon based photovoltaic cells are rather simple and inexpensive, one notable downside of this type of cell is its vulnerability to defect-induced degradation that decreases its conversion efficiency.
In contrast, for more demanding applications such as residential and industrial solar power generation systems, either poly-crystalline or single-crystalline silicon is the choice because of more stringent requirements for better reliability and higher efficiency than the applications in consumer electronics. Photovoltaic cells consisting of poly-crystalline and single-crystalline silicon generally offer the conversion efficiency ranging ˜20% and ˜25% [Zhao J, Wang A, Green M, Ferrazza F. Novel 19.8% efficient ‘honeycomb’ textured multicrystalline and 24.4% monocrystalline silicon solar cell. Applied Physics Letters 1998; 73: 1991-1993.] respectively. As many concerns associated with a steep increase in the amount of the worldwide energy consumption are raised, further development in industrial solar power generation systems has been recognized as a main focus. However, due to high cost ($3 to $5/Watt) of today's Si-based solar cell, this Si-solar cell is not yet widely accepted as an alternative source for the energy solution.
Group II-VI compound semiconductors, for example CdTe and CdS, have been investigated in the context of having industrial solar power generation systems manufactured at a lower cost with maintaining a moderate conversion efficiency, resulted in a comparable conversion efficiency ˜17% [Wu X, Keane J C, Dhere R G, DeHart C, Duda A, Gessert T A, Asher S, Levi D H, Sheldon P. 16.5%-efficient CdS/CdTe polycrystalline thin-film solar cell. Proceedings of the 17th European Photovoltaic Solar Energy Conference, Munich, 22-26 Oct. 2001; 995-1000.] to those for the single crystalline silicon photovoltaic devises, however toxic natures of these materials are of great concerns for environment.
Group I-III-VI compound semiconductors, such as CuInGaSe2, have been also extensively investigated for industrial solar power generation systems. This material can be synthesized potentially at a much lower cost than its counterpart, single crystalline silicon, however conversion efficiency, 18 19%, comparable to that of single crystalline silicon based cells can be obtained, so far, by only combining with the group II-VI compound semiconductor cells [Contreras M A, Egaas B, Ramanathan K, Hiltner J, Swartzlander A, Hasoon F, Noufi R. Progress toward 20% efficiency in Cu(In,Ga)Se polycrystalline thin-film solar cell. Progress in Photovoltaics: Research and Applications 1999; 7: 311-316.], which again raise issues associated with toxic natures of these materials.
A type of photovoltaic cells designed for several exclusive applications such as space, where the main focus is high conversion efficiency and cost is not the main factor. Generally, this solar cell consists of group III-V semiconductors including GaInP and GaAs. Synthesis processes of single crystalline group III-V are in general very costly because of substantial complications involved in epitaxial growth of group III-V single crystalline compound semiconductors. Typical conversion efficiency of group III-V compound semiconductor based photovoltaic cells, as these types of photovoltaic cells are intended to be, can be as high as ˜34% when combined with germanium substrates, another very expensive material [King R R, Fetzer C M, Colter P C, Edmondson K M, Law D C, Stavrides A P, Yoon H, Kinsey G S, Cotal H L, Ermer J H, Sherif R A, Karam N H. Lattice-matched and metamorphic GaInP/GaInAs/Ge concentrator solar cells. Proceedings of the World Conference on Photovoltaic Energy Conversion (WCPEC-3), Osaka, May 2003; to be published.], usually more than 10 times than the conventional Si-solar cell.
All types of photovoltaic cells in the prior arts described above, no matter what materials a cell is made of, essentially falls into one specific type of structure, which usually limits is power generation capability. Usually flat pn-junction structure is used in conventional solar cells (FIG. 1A). Shown in FIG. 1A is a photovoltaic cell comprising a thick p-type semiconductor layer 101 and a thin n-type semiconductor layer 102 formed on an electrically conductive substrate 100. A pn-junction 103 is formed at the interface between the p-type semiconductor layer 101 and the n-type semiconductor layer 102. Incident light 104 entering the cell generate electron-hole pairs after being absorbed by the p- and also n-type semiconductor layers 101 and 102. The incident light generates electrons 105 e and also holes 105 h in the region near the pn-junction 103 and also 106 e and 106 h in the region far from the pn-junction 103. The photo generated electrons (and holes) 105 e and 106 e (hereafter considering only electronics, i.e. minority carriers in p-type semiconductors, and the same explanation is applicable for holes, minority carriers in n-type semiconductors, also) diffusing toward the pn-junction 103 and entering the pn-junction 103 contribute to photovoltaic effect. This is also vice versa for the holes, existing as minority carriers in n-type semiconductor 102. The two key factors that substantially impact the conversion efficiency of this type of photovoltaic cell are photo carrier generation efficiency (PCGE) and photo carrier collection efficiency (PCCE).
The PCGE is the percentage of the number of photons entering a cell and contributing to the generation of photo carriers, which needs to be, ideally, as close as 100%. On the other hand, the PCCE is the percentage of the number of photo-generated electrons 105 e and 106 e reaching the pn-junction 103 and contributing to the generation of photocurrent. For a monochromatic light, the PCGE of ˜100% can be achieved by simply making the p-type layer 101 thicker, however, electrons 106 e generated at the region far away from the pn-junction 103 cannot be collected efficiently due to many adverse recombination processes that prevent photo generated carriers from diffusing into the pn-junction 103, thus the basic structure of current photovoltaic cells has its own limitation on increasing the conversion efficiency. As the minority carrier travel through the semiconductors, the longer the life-time, the less the recombination, which makes the higher the conversion efficiency. Usually, using of thicker and high quality wafer, the conversion efficiency of conventional solar cell can be increased to some extend mentioned earlier. However, this makes the solar cell costly and heavier. In addition to increase the collection efficiency, the absorption of the broad solar spectrum also increase the conversion efficiency.
Furthermore, increasing of the intensity of the solar spectrum helps to also increase the conversion efficiency, thereby increasing the power generation capacity. Conventionally, using of the concentrator separately with solar cell is used to increase the conversion efficiency. It requires additional component with solar cell to concentrate the solar spectrum.
It is highly desirable to have the solar cell structure having (a) high PCCE which is independent to the substrate thickness, (b) the ability of absorption of broad solar spectrum, and (c) self-concentrating capability to increase the intensity of solar spectrum incident per unit area.
FIG. 1B shows typical monochromatic light intensity behavior inside the semiconductor, and the light intensity p at certain depth x can be expressed as p(x)=Poexp(−αx), where Po is the peak intensity at the surface and a is the absorption co-efficient of the semiconductor in which light is entering. As shown in FIG. 1B, the light intensity behavior 108 inside bulk semiconductors is exponential. Carriers (not shown here) generated due to light flux 112 absorbed by pn-junction is only drifted by junction field and can be collected efficiently. Whereas, carriers 106 e and 106 h generated due to absorption of light-flux 110 by semiconductor region 101 are diffused to all direction. Only those which are generated closer (distance equal to or less than the diffusion-length of the semiconductor) to pn-junction, can be collected. Those carriers which are generated far away (distance longer than the diffusion-length of the semiconductor) from pn-junction are recombined and lost. The light flux 112 is usually lost either by going out or absorbed by the substrate. Both PCGE and PCCE are mainly dependent on material and structure of the photovoltaic cells, and today's photovoltaic cells are structured in such a way that (a) wide ranges of solar spectrum cannot be absorbed due to its material limitation, and (b) photo carrier's collection efficiency is lower due to its inherent structure. For example, it is found that typical conversion efficiency of today's crystal-Si based solar cell is ˜18%. Wavelengths of solar spectrum spreads from <0.1 to 3.5 μm in which Si can only absorb ˜0.4 to 0.9 μm of light. ˜50% of light belong to solar spectrum can not be absorbed by Si, due to have its inherent material properties. The rest of 32% are lost due to (i) recombination of photo-generated carriers and (ii) losing of light which is 112 as shown in FIG. 1B, and these two factors are structure dependent. If we could reduce these two factors, ˜50% conversion efficiency can be achieved even in Si-based solar cell. If we could capture different wavelengths of light belonged to solar spectrum by utilizing different material systems or nano-material systems, we could increase the conversion efficiency ideally close to 100%. Furthermore, if the solar cell (photovoltaic cell) detection capability can be extended to infrared-spectrum, then the cell can produce electrical energy not only during day (while sun is present), but also at night (when different infrared is present). Besides, today's solar cell material is not highly radiation-tolerant. In space application, specially, photovoltaic cells should have a structure and material systems, which could generate high-power per unit area and also to highly radiation tolerant. To increase the conversion efficiency (ideally close to 100%), it would be desirable to have photovoltaic cell structures (a) which has larger surface area to volume ratio to capture all the photons (at specific wavelength) entering the cell and a pn-junction that is located at as close to the photo absorption region as possible, (b) amplifying capabilities by concentrating the light incident to its surface, and (c) structure comprising with the material systems having photo responses at different spectrum to efficiently cover a wide range of spectrum of light that enters a photovoltaic cell. It would be further desirable to have solar cell which could generate electric power in both day and night.
In addition to conversion efficiency, cost-effective manufacturing is also another factor requiring some focus. In today's solar cell, high-cost is also one of the main factor in addition to issue of low conversion efficiency. It is found that more than 90% of solar cell is silicon (Si) based solar cell in which crystal silicon (Si) wafer is the based material, and the rest of others are thin-film based solar cell. In manufacturing of Si-based solar cell, more than 50% of cost is originated from Si-wafer cost and the rest are from other cost such as process and integration. Usage of less silicon for the fixed conversion efficiency would help to reduce the power generation cost. It is highly desirable to focus onto the less usage of Si (to reduce the wafer cost) at the same time while increasing the conversion efficiency.